AND GEORGE H. CADY 496 WILLIAMP. GILBREATH
Inorganic Chemistry COSTRIBCTIONFROM THE DEPARTMESTOF CHEXISTRY, ~JN~VRRSITY OF WASHINGTON, SEATTLE, WASHIUCITON
Reactions of the Halogen Fluorosulfates, FOS02F, ClQSO,F, BrOSQ,F, and Br(OSO,F), BY WILLIAM P. GILBREATH
ASD
GEORGE H. CAD\-
Rereived Nouember 15, 1562' T h e new reactive compound, chlorine fluorosulfate, ClOSOZF, was produced by the combination of chlorine with peroxyat about 125' under pressure. It combined with tetrafluoroethylene to form Z-chlorotetradisulfuryl difluoride, S20BF2, fluoroethyl fluorosulfate. This type of behavior was found to be typical of the halogen(1) fluorosulfates of fluorine, chlorine, and bromine. They n-ere added across the carbon-carbon double bond to hirid the halogen atom to one of the carbon atoms and the fluorosulfate group t o the other. Compounds obtained in this manner lyrere: C2FjOS02F, CCl2FCCl20SO21i CFyBrCF20S02F, and perfluorocyclopentyl fluorosulfate. Fluorine fluorosulfate combined with sulfur tetrafluoride t o form SFsOSOsFbut the fluorosulfates of chlorine(1) and of bromine(1) did not react similarly. When used in excess, fluorine fluorosulfate combined with bromine t o give a liquid product having the empirical coniposition Br2.3S03Fz. The same sort of material was formed by mixing equimolar amounts of BrF3 and Br(SOsF)3. It was found also that the latter reactants tended to combine chemically in all proportions when mixed; therefore, the material Br2.3S03F should be regarded as a non-stoichiometric compound
The halogen fluorosulfates, FOSO?FjlBrOS02F,2Br(OS02F)3,zI(OSOsF)3,2and IF3(OSOzF)23 have been characterized, but few of their reactions have been studied. One purpose of the present research was to increase the number of known chemical properties of these substances. Another purpose was to produce and identify a fluorosulfate of chlorine. Previous attempts to do this had been unsu~cessful.~In this research, particular attention has been given to addition reactions of compounds containing the carbon-carbon double bond. This is because fluorine fluorosulfate contains an -OF group Other compounds of this type, specifically CF30F and SF60F, break a t the -OF bond when adding to the carbon-carbon double I t has now been found that fluorine fluorosulfate behaves in the same way and that the other halogen fluorosulfates also break a t the oxygen-halogen bond when adding across the double bond. Bromine was observed to react with fluorine fluorosulfate a t room temperature to form a golden yellow liquid of low volatility. The excess fluorine fluorosulfate could be distilled away, leaving a product having the definite composition Bra.3SOaFr. This formula suggested the possibility that the same material might result from mixing equimolar quantities of BrF3 and Br(S03F)3. When these were mixed, the product appeared to be identical with that formed from Brp and S03Fz. When ratios of BrF3 and B T ( S O ~ Fother ) ~ than 1: 1 were used they also gave a golden yellow liquid of very low volatility, and when BrF3 was used in excess, i t was not possible to remove all of the excess by vacuum pumping. The liquid continued to contain more moles of combined BrF3 than of B T ( S O ~ F ) ~I .t was not found possible to study the system BrF3-Br(S03F)3 by measur(1) F. B. Dudley, G. H. Cady, and D. F. Eggers, J . Ain. C h r m . Sor., 1 8 , 290 (1956). (2) J. E. Roberts and G. H. Cady, i b i d . , 8 2 , 3 5 2 (1960). ( 3 ) J. E. Roberts and G. H. Cady, ibid., 8 2 , 354 (1960). (4) J. M. Shreeve and G. H. Cady, ibid,, 83, 4281 (19G1). ( 5 ) R. S. Porter and G. H. Cady, ibid.,7 9 , 3626 (1957). (6) J. A. C. Allison and G. H. Cady, ibid.,8 1 , 1089 (1959). ( 7 ) Sa M. Williamson and G. I€, Cady, Inava. Chcm.. 1, fi73 (19621,
ing freezing points, because the golden liquids did not freeze upon cooling; instead, they became glassy. A study by n.m.r. was made. however, to learn whether Br2(S03F2)3 would show its own fluorine nuclear resonances while either BrF3 or B T ( S O ~ F when ) ~ , present in excess, would show its typical fluorine resonance. Various mixtures of BrF3with Br(S03F)3were tried a t room temperature but for each mixture only two resonances were observed-one a t the higher magnetic field, for fluorine atoms attached to bromine and one, a t the lower field, for fluorine atoms attached to sulfur. The chemical shifts of the two types of resonance were measured with respect to the fluorine resonance of CCI,F as an external standard. For the chemical shift of fluorine attached to bromine there was a nearly linear increase from 23 p.p.m. for pure BrF3 to 48.9 p.p.m. for a 1 : l ratio of BrF3 to Br(S03F), and from there a nearly linear decrease to 44 p.p.m. for 17.6 mole BrF3. The chemical shift of the fluorine attached to sulfur changed slightly and in a nearly linear manner from -39 p.p.m. for pure Br(S03F)3 to -41 p.p.m. a t 90 mole yo BrF3. Although the maximum chemical shift for one type of fluorine a t 50 mole yo BrF3suggested something special about the composition Brz* 3S03F2,the most significant result of the n.m.r. observations was the implication that rapid exchange occurred involving various molecular or ionic species. Probably polymeric molecules existed in the system, and i t is likely that ions such as BrFp+, Br(SOsF)z+, BrFS03F+, BrF4-, BrF3S03F-, BrF2(S03F)1-,BrF(S03F)3-,and Br(S03F)h- were present. Since BrF3 and Br(S03F)3 appeared to have chemical affinity over the entire range of composition, the yellow liquid of composition B T ~ ( S O ~ Fshould * ) ~ be regarded as a non-stoichiometric compound. An additional bit of evidence about this system was obtained by condensing excess fluorine fluorosulfate upon a sample of dark red bromine(1) fluorosulfate, BrOS02F. When the mixture warmed to room temperature, it became yellow in color. The excess fluorine fluorosulfate was pumped away, leaving a liquid of the empirical compo-
Vol. 2, No. 3, June, 1963
REACTIONS OF
HALOGEN FLUOROSULFATES 497
TABLE I 2 TO 15 p RANGE (Frequencies in cm. -l units)
INFRARED SPECTRA,
CzFsSOsF
CaF603F
CCIzFCClzSOsF
CHzFCHzSOsF?
1500 vvs 1382 w 1249 vvs 1219 s 1158 vs 1093 vvs 1000 w 887 s 843 vs 775 ms
1494 vs 1378 w 1322 s 1288 w 1252 vs 1230 vvs 1155 s 1075 mw 1038 w 990 vvs 849 vs 804 m 697 m
1800 w 1482 vvs 1243 vvs 1121 w 1065 m 984 vs 938 s 896 m 830-852 vvs 741 w
2870 w 1468 vs 1235 vs 1095 m 1036 vs 953 vs 832 vs
CClFzCFzSOaF
1497 vs 1426 m 1261 vs 1221 vs 1185 vs 1153 vs 1123 vvs 985 vs 872 m 843 s 767 m
CFzBrCFzSOaF
2240 w 1498 vs 1315 ms 1256 vs 1217 s 1182 vs 1144 vs 1113 vvs 1080 m 1016 w 945-968 vs 867 ms 843 vs 808 w 766 ms 730 w
chlorine was condensed together with a 0.016-mole portion of sition BrF(S03F)z. The fluorine n.m.r. spectrum was peroxydisulfuryl difluoride, S Z O ~ Fin ~ , a Monel metal reactor composedof twosharp resonances separated by 89 p.p.m. having an external diameter of 6.35 mm. and a volume of 5 ml. with an area ratio of 1:2, the smaller resonance peak A valve on the reactor was closed to confine the materials. This being a t a higher magnetic field than the larger. The vessel then was surrounded by a protective metal shield and was observed chemical shift was the same as that expected maintained a t about 125" for 5 days. The product then was cooled to -78' and subjected to vacuum pumping for 20 hr. for a 1: 2 mixture of BrF3with BT(SO%F)~. to remove the unreacted chlorine. This left behind a material Parts of the BrF3-Br(S03F)3 system probably were which was nearly as pure ClOSOlF as could be obtained by more encountered by Woolfa in 1950. He reported a comcomplicated procedures tried in this study. It was found unpound, BrF2S03F, formed by the reaction of NOSOSF desirable in general t o transfer the substance from one part with BrF2SbF6. He also reported the chemical combiof the system to another because of its high reactivity with traces of water on the glass apparatus. Hydrolysis yielded nation of bromine trifluoride with sulfur trioxide. a red viscous liquid of low volatility. A freshly distilled sample Fluorine fluorosulfate was found to react slowly with was a yellow liquid. Upon standing a t room temperature the sulfur tetrafluoride in a glass vessel a t 70' to produce color gradually deepened and became red. Distillation again the previously known pentafluorosulfur f l u o r ~ s u l f a t e , ~ ~gave ~ ~ a yellow distillate but left behind a red material. SF50S02F,in about 25% yield after a 3-hr. period. Properties of this substance, like those of others mentioned later, are given in Table 11. The analogous reactions using chlorine fluorosulfate or The mass spectrum a t an ionization potential of 70 volts had bromine fluorosulfate with sulfur tetrafluoride in a peaks attributed to Cl+, S02Ff, SO+, SOF+, SO$+, SOz+, and Monel metal reactor a t 80' failed to give the hoped ClSOeF+in order of decreasing magnitude. Peaks for SiFa+, C12+, for addition products. Substances formed by these reand HClt also were present. They probably were formed by the actions included SO2F2, S206F2, SF6, and c12 or Br2. compound reacting with water on glass apparatus. The n.m.r. spectrum, a t 56.4 Mc., had one resonance peak a t a chemical The melting points of the new organo fluorosulfates shift of -33.9 p.p.m. relative t o the external reference, CC13F. produced by adding halogen fluorosulfates to alkenes An impurity peak, probably due to S Z O ~ ForZ S 2 0 5 F 2 , having one generally were not obtained, since, on cooling, glasses fortieth of the area of the principal peak, was present a t a chemirather than crystalline solids formed. Infrared specshift ffrom CClaF) of -47.6 p.p.m.
tra for these substances showed =S/Ostretching
fre-
O\ quencies a t 1468-1500 cm.-l (asymmetric) and 12351261 cm.-l (symmetric), S-F stretching frequency a t 832-850 cm.-l, and various C-F stretching modes between 1000 and 1382 cm.-l. The spectra are given in Table I. With the exception of the material thought to be 2-fluoroethyl flqorosulfate these compounds were thermally stable a t $east up to their normal boiling points. They were not very soluble in water but were dissolved by common organic solvents.
Experimental Chlorine Fluorosulfate, ClOS02F.-A
0.023-mole sample of
(8) A. A. Woolf, J . Chem. SOL.,1053 (1950). (9) H. J. EmelCus and K. J. Packer, ibid,, 771 (1962). (10) G J, Merrill, PhlDx Thesis, University of Washington, 1962,
The reaction of chlorine fluorosulfate with water was violent and yielded oxygen as a product. In one case 24% of the oxidizing equivalent due to chlorine(1) was liberated as 0 2 while the remainder (presumably present as HOC1) was found in solution by an iodometric titration. The proportions of 0 2 and HOC1 formed by hydrolysis varied from one run to another. Chlorine and sulfur were determined in the solution. Reaction of Bromine with Fluorine F1uorosulfate.-A sample of bromine weighing not over 1 g. was condensed in a glass An excess amount of fluorine fluorobulb cooled to -183'. sulfate then was condensed in the bulb. As the temperature of the vessel rose to and remained a t about 25' the substances reacted, giving a golden yellow liquid. The excess of fluorine fluorosulfate then was pumped away while the material was held a t 15'. All of the bromine was present in the yellow product and the latter had a weight indicating that 3 moles of SO SF^ had combined with 1 mole of Br2. Anal. Calcd. for Br2.3SOsFz: Br, 31.0; S, 18.7. Found: Br, 30.6; S,17.0. The hydrolysis reaction used for the analysis was violent and some oxygen was produced. Reaction of Fluorine Fluorosutfate with Iodine(II1) ~ l u o r o s u b
498
WILLIAM
P.GILBREATH AND GEORGE H.C A D Y
Inorganzc Clztmistry TABLE I1
PROPERTIES OF COMPOU\DS CISOaF
M e , "C.
3 i0 3 8 14352 1674 5 - 23 to +43
C?bsS03F
CsFsSOaI:
FCClrSOjI:
7 9805 1505 -23 to +27 22 1 -6 89 23 2 1 714 0 00343 0-20 1 645 218 i: 1 218 1 14 2 14 7
8 1160 1912 -18 to fX7 92 0 -8 75 23 8 1 844 0.00258 3-30 1 i92 330 f 2 330 1 9 6 9 7
7 9332 2194 38-164 161 3 -10 07 23 2 1 834 0 00178 0-30 1 798 282 i 7 284 10 2 11 2 50 1
